Joe Suk

               

I’m a postdoc at Columbia University’s Statistics Department working under Samory Kpotufe, under whom I also received my PhD in 2024. I study multi-armed bandit and reinforcement learning theory, focusing on non-stationary and adversarial problems.

Preprints and Publications

  1. Tracking Most Significant Switches in Infinite-Armed Bandits [code]
    Joe Suk, Jung-hun Kim. International Conference on Machine Learning (ICML) 2025
  2. Adaptive Smooth Non-Stationary Bandits [code]
    Joe Suk. SIAM Journal on Mathematics of Data Science (SIMODS)
  3. Non-Stationary Dueling Bandits Under a Weighted Borda Criterion
    Joe Suk, Arpit Agarwal. Transactions on Machine Learning Research (TMLR) (“Featured Certification”)
  4. When Can We Track Significant Preference Shifts in Dueling Bandits? [code]
    Joe Suk, Arpit Agarwal. Advances in Neural Information Processing Systems (NeurIPS) 2023
  5. Tracking Most Significant Switches in Nonparametric Contextual Bandits
    Joe Suk, Samory Kpotufe. Advances in Neural Information Processing Systems (NeurIPS) 2023
  6. Tracking Most Significant Arm Switches in Bandits
    Joe Suk, Samory Kpotufe. Conference on Learning Theory (COLT) 2022
  7. Self-Tuning Bandits over Unknown Covariate-Shifts [code]
    Joe Suk, Samory Kpotufe. International Conference on Algorithmic Learning Theory (ALT) 2021
  8. Factorizations of k-Nonnegative Matrices
    ($\alpha$–$\beta$) Sunita Chepuri, Neeraja Kulkarni, Joe Suk, Ewin Tang. Journal of Combinatorics
  9. Dihedral Sieving Phenomena
    ($\alpha$–$\beta$) Sujit Rao, Joe Suk. Discrete Mathematics

Misc.